Inhibition of return (IOR) refers to slowed responses to targets presented at the same location as a preceding stimulus. IOR is typically investigated using a cue-target design, in which subjects respond only to the second stimulus of a pair. In such tasks, the measurement of 'true' IOR may be confounded by the effect of non-ocular response inhibition, because the participant must suppress any tendency to respond (e.g. key press) to the first stimulus. This confound may be eliminated using a target-target design, in which responses are made to both stimuli. We assessed the contribution of non-ocular response inhibition to visual IOR, measured in a cue-target task, by testing participants on both cue-target and target-target detection tasks, with identical timings and stimuli. Significant IOR was obtained in both tasks but, at a stimulus onset asynchrony (SOA) of 1400 ms, IOR magnitude was significantly greater in the cue-target condition than in the target-target condition. However, at an SOA of 1800 ms, there was no significant difference in the magnitude of IOR between the two tasks. Thus, a proportion of the total IOR effect observed in visual cue-target tasks can be attributed to non-ocular response inhibition, but this process appears to decay more rapidly than does 'true' IOR, having dissipated by 1800 ms following cue onset.